We cannot prohibit the generation of AI, but we can solve it frictionally (Opinions)
When I write in the Course and Writing Center Coordinator on campus, the instructor asked me how to detect students’ use of Generative AI and how to prevent it. My answer to these two questions is that we cannot.
In fact, it’s getting harder no Use Generate AI. According to a student survey conducted on my campus, some students are even nervous about fear of being tempted to cheat. In the past, students had to look for it, create an account and prompt it. Now the generated AI is integrated into the programs we already use – Word (Copilot), Google Docs (Gemini) and Grammarly, and it’s there to greet us, just like the chocolate I hide in the cupboard every night around 9 o’clock.
Recent grammatical ads emphasize seamless integration of generative AI. Within the first 25 seconds of this Grammarlygo ad, the woman’s confident voice tells us that the grammar is “easy to use” and just “download” and “click button” it is “easy to write better and faster”. The ad also seeks to remove any concerns about the non-humanity and detectability of generating AI: it is “personalization to you”; “know your style, sound, and intentions so that your writing doesn’t sound like a robot”; and it is “customized.” “You can control” and “Grammarlygo helps you become the best version of yourself.” Message: Using Grammarlygo’s generative AI to write is not cheating, but self-improvement.
This ad calls me and we see articles every January of people we want to develop healthy habits. Those who urge us to start working out in the morning urge us to sleep in fitness clothes. If we sleep in clothes, we will reduce the barriers to going to the gym. The focus of some of the most popular self-help suggestions is to reduce friction and allow us to develop the habits we want to build. Like self-help experts, Grammarlygo (and all generated AI companies) strategically seeks to reduce friction by reducing time (“faster), distance (distance) (“writing”) and energy (which is “simple”!).
Where does this leave us? Do we stop assigning writing? Do we assign classroom writing tests? Do we start rating AI-generated jobs by providing feedback from AI?
No.
If we recognize the value of writing is a way of thinking and believe that effective writing needs revision, we will continue to assign writing. Despite a temptation to turn offline, timed writing tests in class, this eliminates the opportunity to practice revision strategies and harms students with learning disabilities as well as English learners.
Instead, like grammar, we can take advantage of self-help advocates and participate in organizational behavior researchers Hayagreeva Rao and Robert I. Sutton called “friction repair.” exist Friction project (St. Martin Press, 2024), they explain how “think and live like a friction fixer, making the right things easier and the wrong things more difficult.” We can’t ban AI, but we can solve it by making generative AI harder to use and making it easier to engage in our writing assignments. This doesn’t mean making our writing assignments easier! The good news is that this approach draws on core practices that have already been effective in writing guidance.
After 25 years working in writing centers at three institutions, I witnessed what stall students are, which rarely lacks motivation. Students using the Writing Center put their work into it, but many people can’t start or get stuck. Here are two ways we can reduce writing assignments:
- Break the research project into steps, including temporary deadlines, meetings, and feedback for you or your peers. Note that feedback does not have to be a comprehensive draft, but can be done in short paragraphs, such as long paragraph project proposals (determine the question, research question and what will be gained if we answer this research question).
- Give students time to start writing projects in class. Have you ever distributed writing assignments, asked, “Is there any problem?” and encountered cricket? If we give students time to start writing in class, we or peers can answer questions that arise, giving students more confidence that they are moving in the right direction and hopefully less towards AI.
Our teachers (inadvertently) have many ways to make our assignment unpleasant: word barriers on the page, lack of blanks, our leadership practices of requirements (citation style, grammatical correctness), the use of SAT words or discipline-specific vocabulary, unskilled vocabulary: All of which can send students any signal that belongs to Getter ot gotter ot gotter ot gotter ot gotter ot gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter new gotter. Sometimes our homework tips even sound annoying because our frustration with past students is misled to current students and manifests itself as a long list of not working hard. The atmosphere is an angry postal note for roommate or partner to stay in the sink… again!
What if we were to reconsider the assignment as a party invitation? When we design party invitations, we have a special goal: we want people to show up, leave their comfort zones and be willing to interact with others. Isn’t this the student we want when we assign writing projects?
If we design writing assignments as invitations rather than assessments, we will make them visually appealing and use passionate language. We will no longer criticize students with all the requirements, but will improve the tempting aspects of the assignment. Find the correctness of APA and MLA formats and syntax, and emphasize the purpose of the job. Transparency in learning and teaching in the framework of higher education is useful for improving assignment layout.
In addition, we can invite students to write real-world audiences and fight with what John C. Bean calls “beautiful questions.” As Bean and Dan Melzer’s Enthusiastic ideas: A professor’s guide to integrating writing, critical thinking and active learning in the classroom (Wiley, 2021) emphasized that the problem is naturally inspiring. From my 25 years of teaching writing experience, students are motivated to write at the time of their motivation:
- Write about their concerns;
- Writing in real genre and real world audiences;
- Share their writing in class;
- Get feedback from its professors and peers on drafts that are based on their own strengths and provide specific tasks for how to improve their work; and
- Understand the usefulness of writing projects related to their future goals.
A three-year study conducted at three institutions confirmed most of these, which asked older people to describe a meaningful writing project. If the assignments are tempting and meaningful, students are more likely to do hard work in learning and writing. In short, we can reduce friction by using language and layout to make them sound appealing and consider by using our audience and designing assignments not only for evaluation, but also for opportunities to explore or communicate.
So, how do we create friction when using Generative AI? As a writing coach, I really believe in the power of writing to figure out what I think and push myself toward new insights. Of course, this is not a new idea. “Writing is really a way of thinking, not just thinking about things that are different, unsolved, mysterious, problematic or sweet,” explains Toni Morrison. If we can really believe this by assigning conventional low-risk writing and strengthening this practice, we can help students see the limitations of outsourcing ideas to generative AI.
With the advent of generative AI, I realized that even though my writing courses were designed to promote thinking about writing, I did not explicitly emphasize the value of writing as a pattern of discovery, so I have rewrite all the free writing tips so I can push this point home: “It’s the low stakes home: “It’s the low stakes sentences, so don’t worry about sentence structure or grammar.” In your speech, please provide the original language or the text you are using to write. Pause and reflect, build new connections, discover new layers of problems, or learn what you don’t know about yourself. “One of my favorite comments is writing an article that is “I’m glad to see your thoughts on the page here.” ”
Furthermore, we can create friction by understanding our students and their writing. We can learn about their writing by collecting ungraded classroom writing at the beginning of the semester. We can get to know our students by canceling classes to hold short one-on-one or group meetings. If we have a strong relationship with our students, it is unlikely that they will intentionally cheat. We can build these bonds by sharing videos about ourselves, writing introduction letters, sharing relevant experiences and failures, writing conversational feedback on student writing, and using alternative grading methods to build these bonds, allowing us to prioritize the process of the above products.
There is no “protective” task, but we can also create friction by assigning writing projects that don’t allow students to rely solely on generated AI, such as Zines, classroom discussions about articles or book chapters or presentations: Generative AI can design slides and write scripts, but can’t introduce material in the classroom. Students are asked to include interactive components in the presentation so that they can interact with the audience. For example, a group of my students gave a speech on Jonathan Haidt’s choice An anxious generationThey asked their peers to check their daily use reports on their phones and responded to anonymous investigations.
Another group created a game that asked the class to guess which books were displayed at one point or another were forbidden. We can assign group projects and give students time to work on these projects in the classroom. It is presumed that if students are responsible for their own groups in some way, there is less chance of abuse of generative AI. We can demonstrate to students by generating AI prompts and asking students to criticize output. This has two beneficial benefits, proving to students that we are savvy while helping them see the limitations of generating AI.
Showing students the limitations of generative AI and the harm it causes will also help create friction. The generated tendency to hallucinate AI makes it a bad tool for research; its confident tone paired with its accuracy, making it the nickname of “Nonsense Machine”. Worse, environmental costs, exploitation of workers, copyright infringement, privacy issues, clear and implicit bias, misunderstanding/proliferation of false information, and more. Students should be asked to study these questions themselves so that they can make informed decisions about how to use Generative AI. Recently, I spent an hour in class, having students study these questions in groups and then introduce what they found in class. Students are particularly subject to privacy violations, environmental impacts, and the use of the work of writers and artists without permission or compensation.
When we focus on seizing students who use Generative AI or prohibit generation, we miss an opportunity to teach students to think critically, and we signal students that we don’t trust them and we will reduce our trust. If we do some friction fixes, we can support students as they work hard to become agile communicators and key users of new technologies.
